For years, there have been several complaints levied against the most divisive of the new-school statistics: wins above replacement.* Detractors cite the "questionable concept" of replacement level, the focus on single-season defensive numbers, or even the estimative quality of WAR as reasons to discard the framework entirely. I believe WAR is hardly a perfect way of estimating value -- but despite these faults, it is currently the best system baseball analysts have to determine a player's overall contribution to a team's success, while placing those contributions on a level where they can compare performance across leagues, positions, and eras.
* - For a more complete breakdown of this framework, check out this.
There's also this: WAR is, really, a framework -- and there are multiple excellent outlets that implement this framework differently in order to try and determine player value. FanGraphs, Baseball-Reference, and Baseball Prospectus, three of the best statistical sites around, all have different metrics that implement the idea of wins above replacement in slightly different ways.
I've always considered this a feature more than a bug, as do many other analysts. These different frameworks use different inputs -- different offensive, defensive, or pitching metrics, etc. -- to determine a player's value. In some cases, these metrics also disagree (perhaps just a little) about the quality of the "replacement level" concept. Regardless, using the term WAR can be confusing, because there's not just one WAR ... there are three commonly used ones: FanGraphs WAR (fWAR), Baseball-Reference WAR (bWAR), and BP's Wins Above Replacement Player (WARP).
I believe in the wisdom of educated crowds, and for that reason, I wanted to try and develop a metric that used all three of these items to create a number that shows the overall view of a players value. As a result, in 2012, I created WAR Index (WARi).
In some ways, WAR Index is a bit of a toy stat -- it does not particularly add anything new to the discussion of these metrics -- but it can be useful in accounting for a player's overall value from multiple, related viewpoints. In the case of pitchers, where the WAR inputs can vary widely (using RA9 instead of FIP as a primary input, depending on the metric), these differences can be both stark and widespread -- but in the case of hitters, big differences can be found between the different implementations as well. WARi gives us a picture of how all three metrics see the player, in a single number.
By scaling the WAR implementations, we also get an ancillary benefit: we can see which systems rate players the most differently, when accounting for the same scale. In most of my WARi spreadsheets, I can sort by which players have the biggest delta between two of the three implementations, and I think this helps us understand what each system values, and why.
So, that's a lot of preface, explaining WAR Index for the folks who haven't checked out this metric since the previous rollout in 2012. WAR implementations have changed a bit in this time. Oh, and if you have an interest in how I actually scale the different WAR implementations, go back to my methodology in the original article. But basically, I use PAs (or IP, in the case of pitchers) and create a coefficient based on the total WAR awarded by a particular system. (Thanks, Tom Tango!)
Blah, blah, blah. On to the good stuff!
2014 WAR Index - Hitters (Google Doc)
So that link above takes you to a copy of my raw WARi spreadsheet that I've uploaded to the cloud. It's always possible that there's a data entry mistake or two in there, so let me know if you spot something fishy. But for now, let's take a look at some leaderboards, shall we?
WARi 2014: The Best of the Best
|Name||Team||PA||rWAR||ADJ fWAR||ADJ WARP||WARi|
|Jose Bautista||Blue Jays||673||6.4||6.39||6.01||6.27|
|Jose Abreu||White Sox||622||5.4||5.38||5.58||5.45|
* - ADJ fWAR and ADJ WARP are the scaled versions of fWAR and WARP.
Check it, the top 25 list includes everyone who managed a WARi of over 5.00 in 2014 -- which by Baseball-Reference WAR (our "baseline WAR") is about equal to an All-Star season. Only two hitters cracked the elusive seven-win barrier: Beyond the Box Score's spirit animal -- Mike Trout -- and
The Mighty Thor Giancarlo Stanton.
To me, there are not a ton of surprises here -- sure, a few guys show dramatic differences from system to system, but it's rare to see a guy who gains or loses more than a win between systems, especially at the top end. Trout, Stanton and Adrian Beltre remain three of the biggest outliers, as their adjusted WARP looks very different than what their fWAR or bWAR represents.
WARi 2014: Please Look Away
|Name||Team||PA||rWAR||ADJ fWAR||ADJ WARP||WARi|
|Allen Craig||- - -||505||-1.0||-1.33||-1.92139||-1.42|
|Kendrys Morales||- - -||401||-1.0||-1.65||-0.85728||-1.17|
|Will Middlebrooks||Red Sox||234||-1.5||-0.77||-1.15674||-1.14|
|Paul Konerko||White Sox||224||-1.1||-1.17||-1.07731||-1.12|
|Raul Ibanez||- - -||280||-1.2||-0.96||-0.88414||-1.02|
|Leury Garcia||White Sox||155||-0.5||-1.18||-1.33122||-1.00|
|Nate Schierholtz||- - -||383||-0.7||-0.95||-1.21830||-0.96|
|Jayson Nix||- - -||91||-0.9||-0.99||-0.87484||-0.92|
Holy crap, Jayson Nix. You really rated almost a full win BELOW replacement in 91 plate appearances? And I don't even want to talk about Justin Maxwell, you guys. Chris Johnson of the Braves is an interesting case here, as he at least managed a positive value via one metric (adjusted fWAR) -- the only person on this list to do so.
Anyways, I feel like this is a good start to the WARi conversation. Hopefully, I'll be able to share the leaderboards for all major league pitchers soon, and we can discuss whatever you feel like: methodology, why some guys were so much better (or worse) by one framework or another, whatever.
But if you ever are interested in what all the major WAR metrics think about a player over the course of a season, all at once, check out WARi. It's a nice way to look at all the elements of value -- from every major statistical perspective -- at once.
* * *
All statistics from FanGraphs and Baseball Reference.
Bryan Grosnick is the Managing Editor of Beyond the Box Score. You can follow him on Twitter at @bgrosnick.